Edinburgh Research Explorer

Duties of Care, , and the Implications of Chester v

Afshar

Citation for published version: Hogg, M 2005, 'Duties of Care, Causation, and the Implications of Chester v Afshar', Edinburgh Law Review, vol. 9, pp. 156-67. https://doi.org/10.3366/elr.2005.9.1.156

Digital Object Identifier (DOI): 10.3366/elr.2005.9.1.156

Link: Link to publication record in Edinburgh Research Explorer

Document Version: Publisher's PDF, also known as Version of record

Published In: Edinburgh Law Review

Publisher Rights Statement: ©Hogg, M. (2005). Duties of Care, Causation, and the Implications of Chester v Afshar. Edinburgh Law Review, 9, 156-67doi: 10.3366/elr.2005.9.1.156

General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights.

Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please contact [email protected] providing details, and we will remove access to the work immediately and investigate your claim.

Download date: 01. Oct. 2021 156 the edinburgh law review Vol 9 2005

Civilian, on the point raised in this case, than the modern codified Civilian systems; both go far to protect the owner’s right through general insistence on the Romanist nemo dat quod non habet. In the modern Civil Law world policy-driven diversity is the norm. A compelling argument for harmonisation is the problem of title laundering by transporting valuable moveables to favourable legal regimes.

D L Carey Miller Professor of Property Law University of Aberdeen

Professor Kenneth Reid most kindly made helpful comments on an earlier version but the author alone is responsible for this paper.

EdinLR Vol 9 pp 156–167 Duties of care, causation, and the implications of Chester v Afshar

A. INTRODUCTION In its decision in Chester v Afshar,1 a 3:2 majority of the House of Lords held that the scope of a doctor’s duty to warn his patient of a non-negligible risk inherent in surgery extends to liability for personal injuries sustained by the patient as a result of the actuation of such risk. Where the warning required by the duty is not given, the patient may claim in damages for the injuries sustained, even although, on normal causal principles, she cannot show that she would not have undertaken the surgery at some later date had the warning been given. The course adopted by their Lordships is one which had already been charted by the High Court of Australia in Chappel v Hart,2 a decision to which the majority of the Judicial Committee made extensive reference. This result may cause consternation in some circles, particularly medical ones, although it is one for which there has been advocacy from leading legal academics.3 Opposition is likely to centre upon concerns about the alleged lack of a causal connection between the doctor’s breach of duty and the injuries suffered by the patient, concerns which were given primacy by the minority. If Miss Chester was unable to prove on the balance of probabilities that she would not have had the surgery on being warned of the risks of the injury (in this case, a condition known as cauda equine syndrome, “CES” for short), then she could not prove that but for the doctor’s failure to warn she would not have been in the same position. I wish to argue that the decision reached in Chester v Afshar is not explicable

1 [2004] All ER (D) 164, [2004] UKHL 41. 2 (1985) 195 CLR 232. 3A Honoré, “Medical non-disclosure, causation and risk: Chappel v Hart” (1999) 7 LJ 1; A Grubb, “Clinical : informed and causation”, (2002) 10 Med L Rev 322; J Stapleton, “Cause-in- fact and the scope of liability for consequences” (2003) 119 LQR 388. Vol 9 2005 analysis 157 primarily by reference to causation at all.4 Rather, a proper explanation of the decision lies in the formulation given by the majority of the scope of the duty of care undertaken by the doctor towards his patient. Because the majority defined the scope of the duty as extending to injuries which were encompassed within the very risk of which the doctor was required to warn, difficulties with the causal connection between breach of the duty and the harm ensuing were thereby overcome at the duty level. The result was that the majority felt it unnecessary to explain the precise causal connection in anything other than the vaguest terms, and certainly without the need to have recourse to any of the traditional tests of causation-in-fact.5 The explanation of the decision thus lies in a normative conclusion of their Lordships (the doctor ought to be liable for this injury) rather than in a causative one (the doctor caused this injury).

B. THE SCOPE OF THE DUTY OF CARE TO WARN OF RISKS INHERENT IN MEDICAL TREATMENT A number of duties of care are imposed upon doctors in relation to their patients, most obviously the duty carefully to diagnose and treat patients’ medical complaints. There is, more to our purposes here, a duty to warn a patient who is contemplating a surgical procedure of any non-negligible risks attendant upon such procedure.6 The question of what foreseeable consequences should fall within the scope of that duty was the issue which the House of Lords had to address in Chester v Afshar.7 In order to answer this enquiry, the majority considered why the duty exists in the first place. Lord Hope, delivering the leading judgment, emphasised the primacy of patient autonomy. The duty to warn exists partly to enable a patient to make up her own mind as to whether or not to have the surgery.8 If, therefore, the duty is broken, one type of loss which properly falls to be considered within the scope of the duty is a diminution of patient autonomy. In Miss Chester’s case this manifested itself in the loss of opportunity to consider, and perhaps take, an alternative course of action (whether not to have the surgery, or to have it performed at a different time or by a different surgeon). Lord Hope also gave a second rationale for the existence of the duty. His Lordship cited approvingly Honoré’s comments, following the decision in Chappel v Hart, that the duty is intended to minimise the risk to the patient.9 Exposure to, or increased exposure to, risk is thus a second type of harm which falls within the scope of the duty. However, Miss Chester’s claim was not expressed in terms either of loss of oppor- tunity or exposure to risk. Her claim was in respect of the personal injuries suffered by

4Cf Lord Bingham, at para 1: “The central question in this appeal is whether the conventional approach to causation in negligence actions should be varied…”; Lord Hope, at para 40: “The issue is essentially one of causation.” 5 The traditional tests of causation-in-fact being the but-for (or sine qua non); the material contribution; and the material increase in risk tests. 6 Sidaway v Board of Governors of the Bethlehem Royal Hospital [1985] AC 871. See, for discussion of this duty in Chester v Afshar, Lord Hope at para 49. 7 This was so despite the insistence of the majority at a number of places that causation was the central issue of the case (see the comments noted at note 4 above). 8 See Lord Hope at para 55. 9 See Lord Hope’s comments at para 80. 158 the edinburgh law review Vol 9 2005 her and consequential losses following therefrom.10 Because of this, their Lordships had to consider whether such loss properly fell within the scope of the doctor’s duty of care. The harm of actual physical injury certainly falls within the scope of the surgeon’s duty to carry out an operation with due skill and care. As we know from the judgment at first instance, affirmed at both stages on appeal, that duty was properly performed by Mr Afshar, so that any injury which resulted from it was not caused by a breach of that duty. However, the fact that this harm falls within the scope of one duty does not necessarily prevent it from additionally falling within the scope of a further duty. The majority in Chester v Afshar conceived that the physical harm was properly to be seen as falling also within the scope of the doctor’s duty to warn. Lord Steyn, in addition to citing the protection of patient autonomy and dignity as within the scope of that duty, stated that the duty “tends to avoid the occurrence of the particular physical injury the risk of which the patient is not prepared to accept”11. This formulation puts the physical injury itself squarely within the scope of the duty. In the same vein, Lord Hope said that “there is no doubt that the injury which Miss Chester sustained when she was operated on by Mr Afshar was within the scope of his duty to warn.” 12 It is not hard to see why their Lordships thought that this should be so. The precise content of the duty of care in question is to warn against the risk of a specific physical injury, so that such physical injury will clearly be in the contemplation of the doctor when considering whether or not to warn the patient.13 The type of injury sustained is thus not in any sense merely coincidental or unrelated to the content of the duty.14 Lord Hope made the further crucial point in relation to the duty that it was “unaffected in its scope by the response which Miss Chester would have given had she been told of these risks.”15 This is of fundamental significance for understanding the determination of the case. What Lord Hope did was to specify that the scope of the duty of care included those hypothetical circumstances which might have eventuated from any response of the claimant to the warning, had it been given by the doctor. These would include both the hypothetical circumstance where Miss Chester responded to a warning of risk by deciding not to proceed with surgery, as well as the opposite hypothetical circumstance where Miss Chester responded by deciding to proceed with surgery. This is important because in so defining the scope of the duty Lord Hope managed to overcome the normal requirement that the claimant must prove that the eventuality complained of would not have happened in any event had the duty been properly performed. Because there was a causal uncertainty about what Miss Chester would have done had she been warned,16 and thus whether she might have undertaken

10 See para 1 of the judgment at first instance. The personal injury, cauda equine syndrome, manifested itself in motor and sensory impairment, as well as neurological impairment. This included very limited mobility, some bowel and bladder disturbance, and continual persistent pain and weakness of the legs: see the judgment at first instance, para 3. 11 At para 18. 12 At para 62. 13 See Lord Hope at para 62. 14 See section C. below on whether the personal injuries sustained could be considered in any sense coincidental to the breach of duty. 15 At para 55. 16 The judge at first instance had been unable to find whether, if the claimant had been properly warned, Vol 9 2005 analysis 159 the surgery with the attendant possibly that she might still have suffered the same physical injury, his Lordship overcame this problem by bringing different counter- factual possibilities within the scope of the duty of care. The very definition of the scope of the duty was held to include different causal outcomes, even if it could not be shown which one would have eventuated had the warning been given by the doctor. It is important to highlight his Lordship’s solution to the problem as lying within an analysis of the scope of the duty of care, and not in the stage of analysis of cause-in-fact. To be sure, his Lordship and his colleagues in the majority could not quite bring themselves to say that the solution lay wholly in defining the scope of the duty. They also said, in the alternative, that the problem was to be resolved by reference to causation. But because the scope of the duty had already been defined in a way which covered the different counterfactual hypothetical outcomes, any problem which might arise at the causation stage was merely a superficial one. Lord Hope could thus say, of the apparent causal problem, that “[o]n policy grounds… I would hold that the test of causation has been satisfied in this case.”17 If, in the minds of the majority, there was a theoretical causal problem, such a problem had practically been nullified through the scope given to the duty of care. The unwillingness of the majority to abandon wholly the notion that the crux of the matter lay in causation, whilst maintaining at the same time that the scope of the duty was the fundamental issue, is understandable, if slightly confusing.18 Whilst it is true that the control mechanisms of liability, including duty of care and causation, ought to be considered separately if we are to avoid conceptual confusion in analysis of the potential liability of wrongdoers, it is also true that that duty of care questions, and thus normative issues, can never wholly be separated from causal questions. Even the simplest duty of care contains an inherent notion of causal connection with the harm in the very expression of the duty. One expresses the doctor’s duty of care as a duty to avoid such actions that the doctor can reasonably foresee will cause harm to the patient. Thus one can never wholly separate normative issues relating to the duty of care from purely mechanical issues of causation-in-fact, although, as in Chester, the normative conclusions about the scope of the duty can render potential causal problems obsolete. The doctor’s duty would not normally extend to: (i) results which are caused through the negligence of another, for instance another medical professional; (ii) results which may merely be coincidental consequences of his or her actions;19 or (iii) results which, due to the inherent uncertainty of how the patient would have behaved, it cannot be demonstrated (using ordinary counterfactual analysis20) would not in any event have

she would with the benefit of further medical advice have given or refused consent to surgery: see Lord Steyn on this point, at para 12. 17 At para 87. 18 Stapleton characterised the decision of the Court of Appeal in the action, which focussed heavily on the question of causation, as being “framed in traditional and wholly inadequate causal language”: see Stapleton, note 3 above, at 420. 19 The notion of coincidental consequences is discussed in section C. below. 20 Counterfactual analysis is the technique used to apply the but-for (sine qua non) test of causation-in-fact, and thus to test whether a factor x (e.g. the negligent conduct of the defendant) was a cause of event y (e.g. the loss suffered by the claimant) in the real world. One postulates a hypothetical past world, 160 the edinburgh law review Vol 9 2005 occurred. However, in order to protect patient autonomy, and to avoid penalising patients for an inability to specify what, under other circumstances, they would have done, the decision of their Lordships ensures that results of type (iii) in this foregoing list are brought within the scope of the doctor’s duty to warn.21 Such a formulation of the duty, like the formulation of any duty of care, is a policy- driven exercise. As Lord Hope succinctly put it: “questions of causation … are better answered by asking whether, all things considered, the defendant should be held liable for the harm which ensued or, on another view, whether the harm was foreseeable as within the risk, or was within the scope of the rule violated by the defendant.”22 This formulation conveys how, for normative reasons, liability for harm which has only an uncertain causal link with the doctor’s duty was none the less brought within the scope of that duty. It is true that the analysis adopted by the majority could have been clearer. The insistence that causation lay at the heart of the problem, when in fact it was at the level of duty, is apt to confuse somewhat.

C. COINCIDENTAL HARM I have characterised the personal injury suffered by the claimant in Chester v Afshar as being a consequence which, due to the uncertainties of human behaviour, had no clear causal connection with the breach of duty. On the other hand, an alternative characterisation of such injury is as a “coincidental consequence” of the doctor’s failure, a concept referred to in the preceding paragraph. Before leaving the question of the scope of the doctor’s duty, I wish to examine whether this concept of coincidental consequence is a better characterisation of the claimant’s injuries. The concept of coincidental harm is not hard to grasp. Where A breaches a duty of care to B, some harm may be caused to B which, while it may not have been unfore- seeable, does not result from the agency of A, in the sense that it was not within the power of A to control or affect the harm, and which is an unlikely occurrence.23 Thus, a taxi driver who, by speeding, brings it about that his passenger is killed by a tree which happens at that precise moment to fall on to the taxi, is not liable for the death of the passenger, as the falling tree is merely a coincidental consequence of his speeding.24

identical to the world as it actually was, save that x was not operative. Instead, one postulates that the

conduct was performed competently, thus x1 (one does not postulate the total absence of x). One then

extrapolates forward from x1 to evaluate what would have happened in this imaginary, counterfactual

world. If the counterfactual outcome, y1, is found to be the same as y (the outcome in the real world, where x was operative), then it is concluded that x was not a cause-in-fact of y in the real world. 21 Although the judge at first instance, His Hon Judge Robert Taylor, does note (at para 82 of his judgment) that, at the assessment of damages stage, regard would have to be taken of any chance that Miss Chester, had she not elected to have surgery, might have gone on to suffer disability as a result of degeneration in her pre-existing condition. 22 At para 85. 23 The Australian judge Mason CJ provided an alternative definition of coincidental consequences as events the risk of which is not increased by the conduct in question: Mason CJ, in March v E & MH Stramore (Pty) Ltd (1991) 171 CLR 506, at 516. Similarly, Stapleton asserts that a coincidental consequence is one “the risk of which is not generally increased by the occurrence of the conduct”, note 3 above at 418. 24 This example is used by Lord Walker in his speech in Chester v Afshar at para 94. Vol 9 2005 analysis 161

Lord Hoffmann offered another example in Banque Bruxelles v Eagle Star Insurance: A mountaineer about to undertake a difficult climb is concerned about the fitness of his knee. He goes to a doctor who negligently makes a superficial examination and pronounces the knee fit. The climber goes on the expedition, which he would not have undertaken if the doctor had told him the true state of his knee. He suffers an injury which is an entirely foreseeable consequence of mountaineering but has nothing to do with his knee.25 Are injuries which result from the eventuation of a risk inherent in surgery, following a doctor’s failure to warn of such risk, properly characterised as a type of coincidental consequence? Lord Walker, in his speech in Chester, suggests not, although the opposite view has been advocated by at least one leading commentator in the field.26 A characterisation of a coincidental consequence as a result which the wrongdoer cannot control or affect suggests that an injury sustained as a result of eventuation of the risk of surgery will not qualify as a coincidental consequence. One cannot assert that it is outwith the power of the doctor to control or affect the outcome, because his warning may affect the decision of the individual patient, and thus may preclude surgery and any adverse effect resulting therefrom. The doctor’s position is thus not analogous to the example of the mountaineer given by Lord Hoffmann, nor of the speeding taxi driver suggested by Lord Walker, for in both those cases the wrongdoer is utterly unable to affect whether the event in question (an avalanche which kills the climber or the tree falling on the passenger) will occur. As the Court of Appeal stated in Chester v Afshar, the “closer analogy is with the mountaineer who consults his doctor because he is afraid that his knee will give way under the strain of mountain climbing, is wrongly assured that it will not, and who is injured because his knee does give way.”27 If one takes the slightly different conception of a coincidental consequence as one the risk of which is not increased by the conduct in question, then the issue hangs on whether it can be asserted conclusively that a doctor’s failure to warn cannot increase the risk of injury. This is discussed in more detail in section D. below, where it is suggested that causal uncertainty means that such a negative causal statement cannot be definitively asserted. Therefore, the doctor’s failure to warn of the risk of surgery cannot qualify as a coincidental consequence characterised in this way. One is led to the conclusion that the injury sustained by Miss Chester was not merely a coincidental consequence of Mr Afshar’s failure to warn. This conclusion adds weight to the permissibility of recovery of damages in the case, as one does not need to overcome the normal presumption that coincidental consequences fall outwith the scope of a duty of care. That does not resolve all difficulties, for, as we have seen, the injuries, if not coincidental, were none the less the result of conduct the counterfactual consequences of which were uncertain due to the unpredictability of the claimant’s hypothetical behaviour. However, as we have also seen, this difficulty was overcome by the definition of the scope of the duty of care as including the consequences of the actuation of the risk, whatever the response of the claimant.

25 [1997] AC 191 at 213D-E. 26 See Stapleton, note 3 above, at 419. 27 At para 44. 162 the edinburgh law review Vol 9 2005

D. CAUSATION Miss Chester was awarded damages in respect of the injury (CES) sustained during surgery. I propose therefore to discuss causation solely in relation to that species of damage, and not in relation to damage conceived of as the loss of an opportunity to seek alternative advice, or as mere exposure to risk. Their Lordships were clearly aware that the but-for test would be problematic for Miss Chester. Indeed they identified that issue, applied through counterfactual hypothesis, as the crux of the causation problem: but for the doctor’s failure to warn, it could not be said that the claimant would not have been exposed to the same risk and thus potentially the same physical injury in any event.28 Their Lordships were also clearly troubled by the view that, in their opinion, the doctor could not even be said to have increased the risk of injury, thus apparently excluding the application of the material increase in risk test. Lord Hope put it: “the failure to warn cannot be said in any way to have increased the risk of injury.”29 This firm assertion that there was no material increase in risk by the doctor was overstating what the evidence could demonstrate. As noted earlier, the judge at first instance was unable to make a finding on what Miss Chester would have done had she been properly warned.30 It may therefore have been the case that had she been so warned, Miss Chester would not subsequently have elected to have surgery. But if so, the surgeon’s failure would indeed have increased the risk of injury. In a hypothetical world where Miss Chester eschews surgery, and where the risk of CES is thus 0, the doctor’s failure to warn has increased the risk of injury from 0 to between 1% and 2% (the stated background risk of CES from surgery even where carefully conducted). Because this possible world may have been the one that would have obtained, it seems to be going too far to say, as Lord Hope does, that the doctor’s negligence certainly did not increase the risk of injury. On the other hand, it is no doubt also correct that the uncertainty about what would have happened had a warning been given means that one cannot reach the opposite positive conclusion, that the absence of the warning did indeed increase the risk of injury. Because of this, although Fairchild v Glenhaven31 is referred to as a case where policy considerations merit a relaxation in the normal rules of causation, it is not cited by their Lordships as a direct analogy with the situation of Miss Chester. Whereas in Fairchild one could assert with certainty that the defendants had each materially increased the risk to Mr Fairchild of injury, such a certainty could not be stated in Miss Chester’s case. The traditional tests for causation therefore do not produce a satisfactory answer in this case, namely one which supports the policies underlying the duty. However, as we

28 See, for instance, Lord Walker at para 90; Lord Hope at para 60. 29 Para 61. Lord Hope continues in the same paragraph: “The risk was inherent in the operation itself.” This is somewhat puzzling, given comments of his Lordship later in his speech. If the failure to warn cannot increase the risk of injury, because the risk is inherent in the operation itself, then presumably, conversely, the presence of a warning cannot reduce the risk of injury, for the same reason that the risk is inherent in the operation. Yet this would seem to contradict Lord Hope’s later approval (at para 80) of Honoré’s view that one of the purposes of the duty to warn is to help minimise the risk to the patient. Either such a warning can have an effect upon the magnitude of risk to a patient, or it cannot, but one cannot consistently hold both views. 30 See note 16 above. Vol 9 2005 analysis 163 have seen, that was not crucial. The solution to this problem had already been located by their Lordships by formulating the duty to warn as encompassing damage actuated by the risk of surgery of which a warning ought to have been given. The majority took the view that the doctor ought to be liable for this sort of injury, a normative choice. To support this normative choice, their Lordships cited the policies underlying the duty (protection of patient autonomy and minimisation of risk to the patient) as sufficient to support a finding that “the test of causation is satisfied in this case”,32 without the need to specify the mechanism by which that was so. In addition to those policy con- siderations, the further considerations of “justice”33 and the legitimate expectations of the public34 were cited by their Lordships as supporting the result. The result of this is that we do not know precisely how their Lordships conceived the nature of the causal connection between the failure to warn and the personal injuries sustained by the claimant. However, it is hard to see how they could have described in more detail the nature of this causal connection. None of the classic formulations of factual causation (whether but-for, material contribution, or material increase in risk) give us a satisfactory answer in this case. Nor would the academically popular “NESS” test,35 first propounded by Hart and Honoré in 1959,36 and later taken up by commentators such as Wright37 and, in a modified form, Stapleton,38 have provided an answer. As Honoré has put it: “We cannot construct INUS39 and NESS conditions for situations in which human beings make and act on decisions.”40 If long

31 [2003] 1 AC 32. 32 Per Lord Hope at para 87. Or, as Lord Hope alternatively puts it, the damage “can be regarded as having been caused, in the legal sense, by the breach of that duty.” It may be queried here whether Lord Hope is using “legal sense” to refer to the concept of legal causation, which as traditionally understood involves normative considerations, or to refer to a specific legal notion (as opposed perhaps to scientific or philosophical notions) of causation-in-fact. 33 See Lord Steyn’s speech, at para 23: “justice and policy demand” the result; also Lord Hope at para 85: “justice requires the normal approach to causation to be modified.” 34 See Lord Steyn’s speech, at para 25. 35 Or Stapleton’s variation of it, the “targeted but-for test”: “Legal cause: cause-in-fact and the scope of liability for consequences” (2001) 54 Vanderbilt LR 941; “Cause-in-fact and the scope of liability for consequences”, note 3 above. In summary, the NESS test formulates a set of conditions operative at the time of the event in question (the injury), and notionally removes factors from that set until one is left with a set of conditions (including the factor under examination, e.g. the defendant’s negligent conduct) which would still have been sufficient to produce the event. If removal of the factor under examination results in a hypothetical situation where the event would not have occurred, it may be concluded that this factor was a cause-in-fact of the event. 36 H L A Hart and T Honoré, Causation in the Law, (1959) (2nd edn, 1985). 37 Mostly recently formulated by R W Wright in “Once more into the bramble bush: duty, causal contribution, and the extent of legal responsibility” (2001) 53 Vanderbilt LR 1071. 38 J Stapleton, “Unpacking ‘causation’”, in P Cane and J Gardner (eds), Relating to Responsibility (2001) 145; also “Legal cause”, note 35 above. 39 The term “INUS condition” was coined by J L Mackie in his seminal causal work, The Cement of the Universe: A Study in Causation, (1974) (2nd edn, 1980). The notion of an INUS condition forms part of Mackie’s theory of the regularity of causes. A NESS condition is a specific instance of an INUS condition, INUS conditions being applicable only to causal generalities. (This footnote was not present in the original of the text quoted.) 40 T Honoré, “Necessary and sufficient conditions in law”, in D G Owen (ed), Philosophical Foundations of Tort Law (1995) 384. 164 the edinburgh law review Vol 9 2005 academic experience has been that a causal connection in a case such as this cannot be shown using established tests, then we should be slow to criticise the House of Lords for its failure to explain the precise nature of the causal connection between Mr Afshar’s failure to warn Miss Chester of the risks inherent in her surgery and the injury she suffered when that risk materialised.

E. WAS THE DECISION REACHED BY THEIR LORDSHIPS CORRECT? As with the decision of the High Court of Australia in Chappel v Hart, a split of 3:2 indicates that the House of Lords did not consider the policy considerations to be uniformly in favour of the claimant. The policy considerations which swayed the majority (Lords Steyn, Hope and Walker) have been considered above. The minority’s five objections may be summarised as follows: (i) The claimant had not been able to show that, on the balance of probabilities, the defendant’s failure had worsened her physical condition. This objection is essentially that the claimant had not satisfied the but-for test of causation.41 (ii) The reasoning of the majority would permit recovery in cases where it could be shown that the claimant would in any event have consented to the operation where properly warned.42 (iii) A modest award of damages might be appropriate in a case like this, but it would be difficult to assess such an award because the risk of injury would differ greatly from case to case. Moreover, the costs of tort litigation would negate the benefit of such modest damages awards. 43 (iv) By extension, the rationale underpinning a decision in favour of the claimant would support a conclusion that a doctor should act as an “insurer” for any damage he or she caused.44 (v) The rationale of the case would not extend to circumstances where a doctor failed to warn a patient, but the surgery was undertaken by someone else. This would be unfair to claimants in such a case.45 These objections are now considered in turn: (i) It is true that the but-for test was not satisfied in this case. However, that test is notoriously inadequate in producing satisfactory results for a number of factual circumstances, such as over-determined outcomes,46 and cases where hypo- thetical past events are indeterminate, either as a result of the unpredictability of human behaviour (as in the case at hand) or the uncertain aetiology of medical conditions (as in Fairchild v Glenhaven). The dismissal of claims in all such cases

41 See Lord Bingham at para 9; Lord Hoffmann at para 32. 42 See Lord Bingham at para 9. 43 See Lord Hoffmann at para 34. 44 See Lord Hoffmann at para 35. 45 See Lord Hoffmann at para 35. 46 One type of “over-determined” outcome is where injury results from the concurrent operation of two separate causes, each of which, on its own, would have been sufficient to produce the injury. An oft- quoted example is of two hunters who, firing simultaneously, each fire a fatal shot at a third party. Vol 9 2005 analysis 165

would be inequitable, and courts have in such circumstances been willing to fashion necessary solutions. In this case, the majority also considered a solution ought to be made available, and achieved this through the formulation of the scope of the duty of care. To object on the ground of failure of the but-for test is therefore to fail to appreciate the underlying rationale of the majority. (ii) At first glance, there is a certain force to this objection. Lord Hope’s formulation of the application of the doctor’s duty, approved by the majority, was that it was “unaffected in its scope by the response which Miss Chester would have given had she been told of these risks”.47 This does seem to suggest that, even where it can be demonstrated that the patient would have elected surgery following a proper warning, she ought still to be able to claim from the doctor if the risk eventuates. This surely goes too far. However, a solution is to suggest that Lord Hope’s formulation of the scope of the duty must implicitly be read as not extending to such a case. After all, his Lordship’s concern was with “the patient who would find the decision [as to whether to elect surgery] difficult.”48 Seen in this light, the ratio of the decision may better be seen as extending only to those cases where either it can be shown that the patient would not have elected surgery following a warning or the patient’s response cannot be shown and she thus takes the benefit of the doubt. (iii) This objection (of Lord Hoffmann) suggests support for the policy of the duty not encompassing the physical injury itself, but merely liability for exposure to the risk of injury, hence justifying only a modest award. This may be because Lord Hoffmann saw the physical injuries suffered as mere coincidental consequences. The force of the objection then rests upon one’s view of what the policies behind the duty require. In this respect, one must simply make a choice as to how the policy concerns ought to be balanced. (iv) It is hard to see what is meant by this objection. The majority’s reasoning clearly proceeds from the view that all damages flowing from a breach of duty are not automatically recoverable, otherwise there would not have been a lengthy discussion of whether the specific harm in question fell within the scope of the duty. There is no suggestion in the majority’s judgments that unforeseeable loss stemming from a failure to warn will be recoverable, or even that all foreseeable loss will be recoverable. (v) It is true that cases where the subsequent surgery was performed by another doctor would not be entirely in point with the decision in Chester. However, there is nothing in the decision of the majority which suggests that the same result could not apply in circumstances where the duty to warn was breached by doctor A but the surgery properly performed by doctor B. Indeed, it seems equitable that the result should extend to such a case. A further concern at imposing this type of liability on doctors who fail to provide a proper warning to their patients might be that it will act as a disincentive for doctors to give advice to patients about treatment options in the first place. The doctor who anticipates that he may inadvertently omit to provide a necessary warning about

47 Lord Hope at para 55. 48 Lord Hope at para 87. 166 the edinburgh law review Vol 9 2005 treatment may choose to avoid all patient questions related to such matters, thus reducing the level of care generally available to patients. This seems rather fanciful. Specialists like Mr Afshar are unlikely to abandon their consultation practices as a result of this decision, and a general practitioner will be unable to exclude the possibility that at some stage a patient may seek advice on treatment. A related concern might be that doctors, faced with a choice of possible treatments, may recommend those which carry lower background risks in order to reduce their potential liability, even if such treatments may not be preferable on clinical grounds. Again, this seems fanciful. A doctor recommending treatment not on clinical grounds but in order to minimise his own potential liability would, by that very act, be failing in his duty to his patient. From the above, it will be evident that the objections to the approach of the majority have not been found convincing. On the contrary, in the light of a weighing up of the competing interests of doctors and patients, the decision in Chester v Afshar represents an acceptable formulation of the scope of liability for the doctor’s duty to warn, and an acceptable outcome on the facts of the case.

F. FUTURE APPLICATION OF THE DECISION The majority did not attempt to set out any rules for the future application of the decision, unlike in other ground-breaking decisions of the House of Lords relating to exposure to the risk of personal injury.49 It is likely that the immediate application of Chester v Afshar will be restricted to cases of medical negligence concerning a failure to warn of the foreseeable risks of treatment where the claimant either would not have elected to have the treatment had she known of the risks, or where it cannot be shown what the claimant would have done had the risks been made apparent. The decision may safely be applied to cases where the failure to warn was that of a party other than the party who, with due care, undertook the treatment. However, it ought not to be applied where the treatment was subsequently performed negligently, and where therefore the physical injury was caused by the botched treatment itself. It remains to be seen whether the decision will be applied in cases other than that of medical negligence, although requests that it be so applied may be expected. However, if that is so, a proper examination of the scope of the relevant duty of care will be required. Each duty of care is supported by its own policy considerations. Those applying to failures to warn by doctors will not necessarily be applicable in other cases, and it will be for future courts to decide whether the policy reasons underlying the scope of a duty to warn of other sorts of risk (for instance, in financial matters) will justify an application of the approach adopted in Chester v Afshar. Consider, for instance, the case of an investment manager who is advising a client as to whether the client should consider entrusting to the manager the investment of a sum of money in a specific financial product. If the manager is aware of specific risks affecting such a product, then he or she ought to give a warning to the client. If a

49 The clearest recent example is Fairchild v Glenhaven, where Lord Rodger, for instance, took considerable trouble in the penultimate paragraph of his speech to lay down what he considered to be the principles which ought to control the application of the material increase in risk test in future cases. Vol 9 2005 analysis 167 warning is not given, the client’s money is invested in the product, and one such risk eventuates, the client may well argue that any losses suffered in result should be recoverable against the manager.50 This argument could conceivably be advanced in a case where the client is not able to say what he or she would have done with the money had the warning been given.51 Ought the Chester v Afshar approach to be adopted in such a scenario? Whilst the facts share many similarities, the policies underlying the scope of the duty imposed upon an investment manager are evidently not identical to those imposed upon a doctor. The nature of the client’s interest—capital speculation— is unlikely to be considered by the courts to be as fundamental, and thus as worthy of protection, as that of health. On the other hand, it is not such a great conceptual jump to substitute “investor autonomy” for “patient autonomy”, or to equate the policy of preventing risk to patients with a policy of preventing risk to investors. In an environment where individuals are being encouraged by Government to take more interest in their own financial planning, especially retirement planning, it can be argued that for the private investor of modest means the interest in protecting financial assets may well rank among his or her most valued. This example suggests that it may not be such a great jump, either in conceptual or policy terms, to extend the rationale underlying Chester v Afshar to cases other than medical negligence. In the absence of guidance from their Lordships, it is a matter of some speculation as to how such arguments might be received by the courts. However, one can speculate with more confidence that, in cases where exposure to risk has yet to produce any physical manifestation of harm, merely the loss of opportunity of avoiding harm, one is dealing with a quite different type of case, a type where the rationale for allowing recovery is much less apparent. One cannot extrapolate from the approach in Chester v Afshar that the same result will be reached in Gregg v Scott,52 a case precisely concerning deprivation of an opportunity of avoiding harm, in which the opinions of the Judicial Committee of the House of Lords are currently awaited.

Martin Hogg Senior Lecturer Edinburgh School of Law

50 Let us say that the risk concerned is not simply the general risk inherent in markets that profits or losses may result, depending upon the general performance of the market, but is a more specific risk involving the type of product, for instance a risk that certain adverse tax consequences might ensue if the Inland Revenue seeks a change in tax treatment of the product, a possibility which the Revenue has made known that it is considering. 51 If the evidence is to the effect that the client would have invested the money in another financial product which would have produced the same level of loss, then the implication of Banque Bruxelles Lambert SA v Eagle Star Insurance Co Ltd [1997] AC 191 is that recovery will not be available. 52 Mr Gregg’s claim is that the defendant doctor’s negligence, in failing to diagnose promptly a cancerous tumour, exposed him to a reduced chance of normal life expectancy. However, he remained in good health at the time of the appeal, so that the harm claimed was for mere loss of opportunity. The decision of the Court of Appeal is reported at [2002] All ER (D) 418.